Learning AI #25

Should AI write your self-evaluation?

Note: See the image included in this post. A picture really is worth a thousand words when we think of the dangers of blind reliance on AI.

Early in her career, writer Suzanna de Baca sat through a performance review that lasted seven minutes. Her manager smiled, said "You're doing a great job," and finished the paperwork before she finished her coffee. No examples. No feedback. Nothing useful.

She spent years wishing someone had taken the time.

The irony is that the answer we're building now is the same kind of nothing, just dressed up better. An AI-written self-review is more polished than "great job." It's probably also more generic than anything Suzanna could have written about her own year.

Most of us (but not all) dread the annual review process where you have to brag about your accomplishments and position yourself for a raise or promotion.

You decide what mattered, what you want credit for, and how you want your manager to see your year of tremendous contributions to the company. That process is, at a minimum, awkward.

Let’s go to a quote from baseball great Derek Jeter. “When you are good at something, other people will talk about you, so you don’t have to.”

You may be the best employee on the planet, but get dinged because you are uncomfortable telling other people how great you are. Something is broken here.

Gemini had the following useful suggestion on how to incorporate AI into your self-evaluation:

How to Do It Right (The "BRAG" Method)

To get a genuinely great self-evaluation, don't just type "write a self-evaluation for a healthcare professional." Instead, give it raw ingredients using this framework:

1. B - Background: Tell the AI your role and the specific competencies your company evaluates (e.g., leadership, technical execution, communication).

2. R - Raw Notes: Feed it a messy, bulleted list of things you actually did this year. Include projects, problems you solved, and any metrics you remember.

3. A - Adjust the Tone: Explicitly instruct the AI on how to sound. For example: "Write this in a professional, objective, and confident tone. Avoid overly flowery language or corporate clichés."

4. G - Grounding: Remind it to stick strictly to the facts you provided.

Example Prompt to Copy and Paste:

The version that will derail your career? Paste your job description and last year's goals into an AI, generate 400 words, fix the pronouns, and submit it. Your manager doesn't know the difference, but you do. And next year, when you sit across from someone and try to explain what you're worth, you'll have practiced that conversation with a chatbot instead of yourself.

A seven-minute review where your manager says nothing useful is one failure. Letting a generic AI write your own year-in-review teaches you nothing and starts to build a fragile foundation for a career. It is OK to use AI in creating your self-evaluation, just do it right.

Things I think about

Your skin is completely replaced every 27 days. You'll have about 1,000 different skins in your lifetime.

Check out the 10-minute interview I did on a podcast.

Free gift for my subscribers

Here is the link for a free audiobook chapter from my new book, No Adult Left Behind. AI is Learning. Are You?

There is no login or anything like that, just click and listen. The chapter is all about how to use prompts with AI. A good starting point for anyone who is serious about learning AI. (The audiobook is narrated by my AI-cloned voice.)

Response from Gemini

Using AI to write your self-evaluation is a highly effective strategy, provided you treat it as a collaborative co-pilot rather than letting it run on autopilot.

Writing about your own achievements can feel awkward or time-consuming, and AI is exceptionally good at translating casual notes into professional, polished corporate language. However, relying on it completely carries some risks.

Here is a breakdown of how to approach it safely and effectively:

The Benefits: Why You Should Use It

• Overcomes Blank-Page Syndrome: It’s much easier to edit a draft than to start from scratch.

• Polishes Tone: AI can instantly convert a bullet point like "Fixed a bunch of broken data pipelines" into "Streamlined critical data infrastructure, reducing system downtime and improving processing efficiency."

• Removes Implicit Bias: Studies show people often downplay their own achievements in self-evaluations. AI doesn't feel awkward about making you sound impressive; it just looks at the data you give it.

The Risks: What to Watch Out For

• The "AI Voice" Trap: Standard AI outputs can sound overly generic, robotic, or hyper-inflated (using words like delve, testament, revolutionize, or spearheaded five times in one paragraph). Your manager knows how you talk and write; if the tone is suddenly radically different, it will stand out.

• Data Privacy & Compliance: Never paste proprietary company data, patient records, unannounced financial figures, or sensitive client names into a public AI model. Keep your prompts focused on high-level achievements and metrics.

• Hallucinations: AI won't remember your year accurately unless you give it the details. If you ask it to blindly guess what you did, it might make up projects or metrics you never actually touched.

How to Do It Right (The "BRAG" Method)

To get a genuinely great self-evaluation, don't just type "write a self-evaluation for a healthcare professional." Instead, give it raw ingredients using this framework:

1. B - Background: Tell the AI your role and the specific competencies your company evaluates (e.g., leadership, technical execution, communication).

2. R - Raw Notes: Feed it a messy, bulleted list of things you actually did this year. Include projects, problems you solved, and any metrics you remember.

3. A - Adjust the Tone: Explicitly instruct the AI on how to sound. For example: "Write this in a professional, objective, and confident tone. Avoid overly flowery language or corporate clichés."

4. G - Grounding: Remind it to stick strictly to the facts you provided.

Example Prompt to Copy and Paste:

The Golden Rule

Always edit the final output. Use the AI to generate the 80% baseline, then spend 10 minutes injecting your specific voice, tweaking the metrics, and ensuring it genuinely reflects your contribution.

What industry or specific role are you writing this evaluation for? I can help you structure a custom prompt or polish your raw bullet points if you want to brainstorm them here.

Response from ChatGPT

Yes. In most workplaces, it's perfectly appropriate to use AI to help write a self-evaluation—as long as the content is truthful and reflects your own work.

A good way to think about it is:

• Fine: Using AI to organize your thoughts, improve grammar, make your writing more concise, or turn bullet points into polished prose.

• Fine: Asking AI to help you highlight accomplishments or frame challenges constructively.

• Not fine: Having AI invent accomplishments, exaggerate your impact, or claim you did work that you didn't do.

• Potentially an issue: If your employer has a policy that prohibits using AI for internal documents. Some companies (particularly in finance, healthcare, government, and defense) have rules about entering confidential information into public AI tools.

The strongest self-evaluations usually combine your own facts with AI's writing assistance. For example, you provide:

• Projects you completed

• Quantifiable results (revenue, cost savings, customer satisfaction, etc.)

• Challenges you overcame

• Goals for next year

Then AI can turn those into a clear, professional evaluation.

For example, instead of writing:

AI might help you rewrite it as:

That's polishing your ideas—not replacing them.

If you want, I can also help you write a self-evaluation that sounds authentic rather than obviously AI-generated. In my experience, those tend to be more effective with managers.

**********